Refactoring ADDA rectangular dipoles part according to article

[dsmunev]

1. all description, including references to equations, should be relative to (Smunev et al. 2015) - since it contains some improvements and is easier to check.
2. the rect_dip part should be integrated with the rest (i.e. first cycle over formulations and then if(isUseRect) inside each, and incompatibility statements, if necessary).
3. Preliminary calculations (or indexing of tables) should be done in a separate function (like CC_Init) - this will become especially important in the future - Continuous polarizability  #12
4. This separate functions may use or define all constants from const.h (like LDR_B1, SO_B1,..). We know this constants only for some special values, like 1:1:1 but can compute for others (at least, for some formulations). Maybe the predefined constants can be moved/removed from const.h.
5. Related to the above. Current constants c1-c3 should be expressed through LDR_ ones, or vice versa. I.e. there should be only one set of hard-wired constants.

Related issues:

#245
#196

[myurkin]

Also related to #251

[myurkin]

When doing this, try to make rectangular dipoles compatible with anisotropy (-anisotr)